Security flaws surface across AI, smartphones and telecoms

It’s been a busy stretch for tech security and infrastructure. First, OpenAI is dealing with a serious privacy lapse after an investigation showed that external contractors read user chats through something called Project Lily. The company says this amounts to a data breach. At the same time, OpenAI also reported a user who made threats in ChatGPT to the FBI. On the device side, Google confirmed that attackers are exploiting a modem flaw to compromise some Pixel smartphones, though they didn’t say how many. Cisco, meanwhile, has rushed out an emergency patch for a zero-day in its Identity Services Engine that lets hackers bypass login controls entirely. Beyond security, there’s progress elsewhere: the US is funding Vodafone Ukraine to upgrade its telecom network and bring in 5G, and Fitbase has added a chat function to its app so fitness club members can talk inside the platform. These events show how privacy, device safety and connectivity are all moving at once — sometimes in troubling ways, sometimes as part of broader upgrades.